Neuronal networks in the spinal cord of limbed vertebrates generate motor rhythms that usually include alternation between activation of hip flexor and hip extensor motoneurons. We study these rhythms in a spinal turtle with complete spinal cord transection just posterior to the forelimb enlargement. We examine spinal cord circuitry responsible for the production of 3 motor strategies, "forms", of scratching (rostral, pocket, caudal) and 2 forms of swimming (forward swim, backpaddle). We test the "bilateral shred core" hypothesis: spinal cord circuitry involved in the production of a specific rhythmic behavior in a hindlimb is bilaterally distributed and is responsible for the production of more than one rhythmic behavior. We developed a spinal preparation with transverse hemisection anterior to the hindlimb enlargement that produces hip flexor rhythms in the absence of hip extensor activity in response to ipsilateral stimulation in the mid-body rostral scratch receptive field (J Neurosci 18:467-479, 1998). This establishes that hip flexor circuits can be rhythmogenic in the absence of hip extensor circuit activation. This preparation also generates "reconstructed" normal rostral scratching with hip flexor/extensor rhythmic alternation in response to 2-site stimulation with one site in the intact-side rostral scratch receptive field and the other site in another scratch receptive field. These experiments support the bilateral shared core hypothesis for the 3 forms of the scratch. We propose experiments that study scratch motor patterns in the spinal immobilized turtle with transverse hemisection. We analyze this motor patterns along with synaptic drive in motoneurons and activity patterns of descending propriospinal interneurons. We test specific predictions of the bilateral shared core hypothesis, e.g., we examine post-synaptic potentials in contralateral hip motoneurons during scratching motor rhythms in response to tactile stimulation in an ipsilateral scratch receptive field. We also propose experiments that study scratching and swimming in a spinal turtle with movement. We measure muscle activity patterns and hindlimb kinematics. We test a version of the bilateral shared core hypothesis stating that neural elements that produce scratch rhythms also produce swim rhythms. Turtle spinal cord is similar to that of other vertebrates, including humans. The spinal mechanisms that we reveal in turtle serve as working hypotheses for studies of motor rhythm generation in other vertebrates, including humans.